BSDSUCCESS STORIES 11 Success stories with BSD Adventures in BSD How BSD Keeps Me Sane FreeBSD at Shannon Medical Center BSD in a Panic You Haven’t Had E-mail Since When? FreeBSD saves a dot-org, and maybe me, too! A FreeBSD Success Story (and DragonflyBSD too) BSD In a Microsoft Office Museum Guardian OpenBSD Saves the Day A FreeBSD Implementation Open Source Software in Co-operation Ireland Introduction The BSDs have earned a reputation for stability, security, and performance. That's no accident. Their Unix heritage provides the power. Careful attention to detail and correctness on the part of their developers provides the finesse. This has produced highly-reliable systems that are easy to administer, flexible to any task, and compatible with tens of thousands of pieces of open source software. No wonder BSD is a popular choice with ISPs and the basis for Apple's Darwin! The BSDs are also proving themselves popular with both sysadmins and users looking for alternatives to expensive, proprietary, and inflexible operating systems. This collection of BSD Success Stories—written by ordinary users-- shows just a few of the many ways in which the BSDs solve real problems all day, every day. Adventures in BSD by Michael Josefsson I am sometimes asked “Why do you use old LAN software to run under Windows. FreeBSD?” My usual ‘auto-response’ is I don’t remember why, perhaps I was a “Because it is the best!” While I truly think victim of my own ignorance. After reading it is the best, this reply does not really the Windows docs I managed to set up a answer the question. I hope this article common workgroup for the two labs. With will better explain my feelings. a 486/66 Windows box working as a file Way back in 1994 I became the manager of server we again had a fully functioning a hardware laboratory at my old University. lab.or so we thought! I was to manage the labs in all aspects, from I found that it was possible to access the computers down to data sheets, pliers, and server’s shared area and was quite pleased students. With over 300 students attending with my doings. I compiled some typical each year, this was a very time-consuming code from a client machine and everything task. Of the three labs, only two had ran as expected. I did not test network computers and networks in them. In the performance under a higher load. The third lab, 16 serial lines were connected idea never crossed my mind. through a Xyplex terminal server to a During one compilation, several large, computer of unknown pedigree. (It later multi-megabyte files were created in the turned out to be a Sun 3.) I had been into project’s home directory. This was in itself computers since the early 80s and PCs since not a problem as the network’s bandwidth 1989, but I had never touched a network was not saturated. However, all this data of any kind. moving in and out of the server turned out The two labs were connected by a Lantastic to be detrimental to its health. With 10-15 LAN. Only one area of the server’s disk was clients chewing away, the load on the server shared and every student stored his own increased. After a period of time ranging home-directory on it. As long as everything from a few minutes to several hours, the worked fine, and it generally did, I had server would unexpectedly crash. A reboot better things to do than trying to under- got everything working again. All source stand what this network setup was all files had been auto-saved before entering the about. The odd downtime was mostly due compilation stage so it was easy to start over to students tampering with the RG-58 coax and hope for more success the second time. that ran from one machine to another Still, the situation was less than satisfactory. continuing on into the next lab. Problems At this point, OS/2 entered the story. In were easy to locate. I knew the coax had the outside world, Windows 95 was being to be terminated with a 50-ohm BNC deployed. I got my hands on an old OS/2 terminator and moving this terminator 2.1 demo CD. After several problems I got to strategic places along the cable always it up and running. I thought it was cool led me to the culprit. and ordered my own OS/2 Warp soon after. Then there was a software upgrade. The Warp was my favourite desktop for years labs had been running MS-DOS with a to come. I really liked the smooth user DOS-based compiler for many years. The interface and enjoyed the increased stability new version of the compiler required MS compared to the Windows version I had Windows. (At the time this meant Windows used earlier. A co-worker with a similar for Workgroups 3.11.) I could not get the interest in computers used Windows 95 2 and had to re-install every three to four FreeBSD. In the beginning of August, when months. The system would somehow clog I had to wire-up the lab’s new infrastructure, itself up beyond recognition. My OS/2 I made FreeBSD run the lab for me. That machine ran happily all the time. Great! turned out to be a wise move; not one Except for the RSA DES Challenge. Being spontaneous reboot occurred during the outside of the United States I took an years to come. interest in the Swedish-run SolNet DES- With a busy lab, reliability is of major impor- attack. With SolNet’s limited resources tance. The logs show students accessing the there was quite some time before an OS/2 server around the clock. FreeBSD really is executable was published. The situation reliable; the server’s longest uptime to date annoyed me. is 220 days. So what to do? I was unwilling to give-up By now, things were moving fast. The third OS/2 and surrender to Windows 95. I had lab was also computerized. (The introduction started to experiment with Linux, a Unix- of Microchip’s PIC line of processors neces- like system but felt very little enthusiasm for sitated this.) I added a couple of hubs and it. Because of an earlier experience with HP- the labs were interconnected. The next UX, I was under the impression that Unix summer I set up a local name server and only represented an extremely complicated a private domain on a FreeBSD box with way of doing things, and therefore, Unix two NICs—one for the internal net and one was ruled out. for external access. There was no forwarding At this time, there were executables for between the two interfaces. This setup FreeBSD available. Not knowing much allowed me, as admin, to download about FreeBSD, I made an FTP install software which I then installed on the of 2.1.7-STABLE. The DES-client ran as clients. Each project team in the lab had expected. I figured out nice commands; its own home, with permissions fixed so kill -STOP, kill -CONT and that putting they could not peek at each other’s work. an ampersand (&) after the command line They could also edit files at home and ran a job in the background. Cool! I could upload via FTP. I added the system user manage every aspect of the program. A taste as a way to keep a central repository of of a new world! This was very enticing. programs and files that every client would My success was short-lived. There was need—Adobe Acrobat for Windows for example, soon a new DES client requiring newer libs, printer drivers, upgrades and other stuff. which forced me to install 2.2.2. This time Until now all data sheets had been in a I ordered the 2 disc FreeBSD set from folder in my room. The projects’ supervisors Walnut Creek sometime around June 1997. had access to the folder and make copies Before the summer I had plans to replace which they handed out to the project teams. the lab server with OS/2, but during the There were 95 steps to the photo copier, summer, I experimented with FreeBSD and with 95 steps back from it, handling and learned about TCP/IP, Samba, FTP the data sheets became more and more and Unix. cumbersome as a variety of components The client machines in the lab were using were added. What to do? Install Apache! Windows 95 and since we had to get rid Putting the data sheets on the server in of Windows for Workgroups anyway, I PDF-format has eliminated the actual paper installed FreeBSD on each and every handling. This was a truly brilliant move machine. I made further experiments with that I wish I had thought of earlier. Now the r*-services (rsh, rcp, ruptime, etc.) each project group could peruse the among other things. All in all, the experi- available components’ data sheets, both from ments made me confident with managing home and when in the lab before committing 3 Adventures in BSD, continued the device into their project. Some students with vinum(8). Thus the entire /usr is printed out data sheets of course, but I see a safeguarded from disk failure. To ease new trend of more and more students reading the installations and administration, each client specs directly off the screen. has a P:\ (P for program) directory with all There is a lot of activity in the labs.